Jump to content

sgl_imaging_challenge_2021_annual.thumb.jpg.3fc34f695a81b16210333189a3162ac7.jpg

Which Sensor?


Recommended Posts

You might be oversampling with the 183, 533 looks a bit more forgiving. Thats with no reducer. https://astronomy.tools/calculators/ccd_suitability

Worth also checking out the field of view calculator on there for your intended targets :) 

I have no experience with these particular sensors though. You'll notice coming from CCD to CMOS (as I am, I'm still getting to grips with the ones in my sig.) there's a few things to consider with amp glow, bit depth & calibration in comparison to CCD. However, the sensitivity is quite something! 

Link to comment
Share on other sites

Posted (edited)
1 hour ago, Sp@ce_d said:

You might be oversampling with the 183, 533 looks a bit more forgiving. Thats with no reducer. https://astronomy.tools/calculators/ccd_suitability

Worth also checking out the field of view calculator on there for your intended targets :) 

I have no experience with these particular sensors though. You'll notice coming from CCD to CMOS (as I am, I'm still getting to grips with the ones in my sig.) there's a few things to consider with amp glow, bit depth & calibration in comparison to CCD. However, the sensitivity is quite something! 

Going on what Olly was saying earlier, the 2.4um pixels of the IMX183, would need very accurate guiding, although it could be binned at 2x2, that effectively brings the 20mp Sensor down to 10mp. The IMX533 has 3.7um pixels should, as you say, be a bit more forgiving on the guiding, without the needs to Bin,  and at 9mp it would on a par with a 2x2 binned 183. I think I'm heading in the direction of a IMX533 based camera. I also like the square format, as I had with Kodiak KA14022 in the SXVR H16.  As such considering either ZWO or Altair equivalent. Same 'beast' different badge.

The 183 would be giving a resolution of just 0.6" per pixel, which is getting close to oversampling, whereas the 533 would give 0.96" per pixel, which is close to 1".

Edited by centroid
Link to comment
Share on other sites

44 minutes ago, centroid said:

The 183 would be giving a resolution of just 0.6" per pixel, which is getting close to oversampling, whereas the 533 would give 0.96" per pixel, which is close to 1".

How about ASI294mc?

You'll be oversampling with 1"/px as well. In reality with that setup - you want to target 1.6-1.8"/px resolution.

Add Riccardi x0.75 FF/FR and you'll be "right on target" with ASI294mc at 1.6"/px

Link to comment
Share on other sites

The IMX 533 the more modern CMOS chip with the magical HCG switch at 100gain and zero amp glow (which isn't hype).  As always, Vlaiv is spot on re sampling rates.

Link to comment
Share on other sites

26 minutes ago, MartinB said:

The IMX 533 the more modern CMOS chip with the magical HCG switch at 100gain and zero amp glow (which isn't hype).  As always, Vlaiv is spot on re sampling rates.

I'm don't think that the ASI294mc is cooled, which would concern me somewhat. I am very much leaning towards the IMX 533 at the moment. Its just the choice of scope that is the conundrum at the moment. Don't really want to go below 100mm, and as I was so impressed with the WO FLT 110, is the reason for leaning towards the Altair 115 Triplet. at £1200, which is close to the £1k budget that I have in mind. Decisions, decisions, decisions 😅

Link to comment
Share on other sites

1 hour ago, centroid said:

I'm don't think that the ASI294mc is cooled, which would concern me somewhat.

There are two versions, the ASI294MC with no cooling and the ASI294MC-Pro with TEC cooling down to -40°C.

I don't know what your budget is for the camera, but I thought I'd throw the ASI2600MC-Pro into the mix with the larger Sony IMX571 APS-C sensor. :D

  • Thanks 1
Link to comment
Share on other sites

7 hours ago, Budgie1 said:

There are two versions, the ASI294MC with no cooling and the ASI294MC-Pro with TEC cooling down to -40°C.

I don't know what your budget is for the camera, but I thought I'd throw the ASI2600MC-Pro into the mix with the larger Sony IMX571 APS-C sensor. :D

I'm looking at at all options at the moment, with a camera budget of circa £1k. 'Back in the day' (pre 2014), I was into astro-imaging in big way, built an observatory with warm room, and had lots of gear. Now its just going to be just a tripod mounted scope, in the garden, and camera, to have a 'dabble' again', and scratch an itch, that has never gone away  😅. With the mount, scope, and camera, its going to be around £3k, which is plenty enough this time around.

Link to comment
Share on other sites

Ok, thanks to Vlaiv and Budgie, I think I have narrowed the choice down to two cameras. The ZWO with the IMX 533 sensor, or the ZWO with the IMX 294 sensor. The 533 with an f/7 refractors would give a res of 0.96" per pixel, and the 294 1.19" per pixel.

The 533 claims to have zero amp glow, supported by Martin, whereas the 294  makes no reference to it. There is only £100 difference between the two.

The 294 is APS-C size, which raises the question, would this require a Field Falttener if used with an 115mm f/7 refractor?

The input and advice from you guys has bee very helpful, and is much appreciated.

I should add that while back in the day, I always used Starlight Xpress cameras, this time around, they, along with Atik, are way beyond the budget allocated.

Link to comment
Share on other sites

17 minutes ago, centroid said:

The 294 is APS-C size, which raises the question, would this require a Field Falttener if used with an 115mm f/7 refractor?

It is actually closer to 4/3 being 23mm diagonal sensor.

I would strongly suggest you consider adding FF/FR to the setup. Not so much for field flattening effect as for focal length reduction (get field flattener that is also reducer).

 

  • Thanks 1
Link to comment
Share on other sites

39 minutes ago, centroid said:

The 533 claims to have zero amp glow, supported by Martin, whereas the 294  makes no reference to it. There is only £100 difference between the two.

The ASI294MC-Pro does have amp glow showing on the subs but this is removed easily with the use of Dark calibration frames. I did have problems with some residual amp glow in the final stacked image but removing the Bias calibration frames from the stacking process sorted that and I don't use Bias frames now. ;) 

I've also read that some have had issues with producing good Flats with this camera but I found that flats of over 1.5 seconds exposure & combined with Dark-Flats at ADU of 250000 works fine. 

Link to comment
Share on other sites

1 hour ago, vlaiv said:

It is actually closer to 4/3 being 23mm diagonal sensor.

I would strongly suggest you consider adding FF/FR to the setup. Not so much for field flattening effect as for focal length reduction (get field flattener that is also reducer).

 

I will look for any field curvature once I use the scope, and if present, will obtain a FF. If I go for the IMX533, which will give me just under 1" per pix res, the mount handles that, and the image looks good, then fine. If not,  I will add a FR.

Link to comment
Share on other sites

4 minutes ago, centroid said:

I will look for any field curvature once I use the scope, and if present, will obtain a FF. If I go for the IMX533, which will give me just under 1" per pix res, the mount handles that, and the image looks good, then fine. If not,  I will add a FR.

Image can look good and mount can handle well even at crazy oversampling - that is not really the point.

image.png.871ceac4776cfd9fd8c756b426ee410e.png

This image was captured at 0.5"/px. It looks fine enough, right?

Point with over sampling is spreading of the light over more pixels than necessary. If above image is capture at 0.5"/px then light coming from 1"x1" is spread over 4 pixels - each pixel gets only 25% of that light - so signal per pixel is much smaller and SNR is much worse than it could be.

Again to image above - that image has resolution of about 2"/px. I binned it to 1"/px because it made no sense to keep it at 0.5"/px - in fact I consciously made decision when I was putting my gear together that I'll be using binning with that particular scope as it is 8" RC scope with 1600mm of FL.

Here it is at 1"/px viewed at 100% zoom:

image.png.80f49497f3544c8870f23100057438e1.png

You can clearly see that it is no longer as sharp as it can be - simply resolution is not there for 1"/px (and that is with 8" scope).

In general, for DSO imaging, resolution of your image will depend on three factors - seeing, mount performance (guiding RMS) and scope aperture. You can very easily measure what is proper resolution for any given image - simply by looking at FWHM of stars in the image. Sampling rate should be FWHM in arc seconds divided with 1.6.

Problem with smaller scopes is that their Airy disk alone is getting bigger and when you add seeing and guiding on top of that - you won't be able to achieve very fine resolution. In fact - having image that actually has resolution of 1"/px is very very hard and only really achievable with 8"+ apertures in best conditions.

Just for comparison - Diameter of airy disk for 115mm telescope is 2.28". In order to sample at 1"/px with such scope you need 1.2" FWHM seeing and no mount guiding errors.

HEQ5 that you plan to use is very decent mount - I have it as well, however, it will guide only down to 1" RMS out of the box. You need to seriously mod it and tune it to get it to guide at 0.5" RMS. With 1" guide RMS - you won't be able to achieve 1"/px with 115mm scope - even if sky is completely still and seeing is perfect (no seeing error). Combined blur from scope aperture and mount tracking errors will be greater than 1.6" FWHM stars.

If you are so set at 533 sensor - by all means go for it. And if you don't want to get reducer flattener as well (533 won't need it as it is 1" sensor), that is fine as well. In that case, do seriously consider how you are going to process your data. You'll need to bin x2 at least (and binning of OSC data is not straight forward as with mono - if you debayer with interpolation and then bin - you don't get same SNR improvement as most of your pixels have correlation).

  • Thanks 1
Link to comment
Share on other sites

2 minutes ago, vlaiv said:

Image can look good and mount can handle well even at crazy oversampling - that is not really the point.

image.png.871ceac4776cfd9fd8c756b426ee410e.png

This image was captured at 0.5"/px. It looks fine enough, right?

Point with over sampling is spreading of the light over more pixels than necessary. If above image is capture at 0.5"/px then light coming from 1"x1" is spread over 4 pixels - each pixel gets only 25% of that light - so signal per pixel is much smaller and SNR is much worse than it could be.

Again to image above - that image has resolution of about 2"/px. I binned it to 1"/px because it made no sense to keep it at 0.5"/px - in fact I consciously made decision when I was putting my gear together that I'll be using binning with that particular scope as it is 8" RC scope with 1600mm of FL.

Here it is at 1"/px viewed at 100% zoom:

image.png.80f49497f3544c8870f23100057438e1.png

You can clearly see that it is no longer as sharp as it can be - simply resolution is not there for 1"/px (and that is with 8" scope).

In general, for DSO imaging, resolution of your image will depend on three factors - seeing, mount performance (guiding RMS) and scope aperture. You can very easily measure what is proper resolution for any given image - simply by looking at FWHM of stars in the image. Sampling rate should be FWHM in arc seconds divided with 1.6.

Problem with smaller scopes is that their Airy disk alone is getting bigger and when you add seeing and guiding on top of that - you won't be able to achieve very fine resolution. In fact - having image that actually has resolution of 1"/px is very very hard and only really achievable with 8"+ apertures in best conditions.

Just for comparison - Diameter of airy disk for 115mm telescope is 2.28". In order to sample at 1"/px with such scope you need 1.2" FWHM seeing and no mount guiding errors.

HEQ5 that you plan to use is very decent mount - I have it as well, however, it will guide only down to 1" RMS out of the box. You need to seriously mod it and tune it to get it to guide at 0.5" RMS. With 1" guide RMS - you won't be able to achieve 1"/px with 115mm scope - even if sky is completely still and seeing is perfect (no seeing error). Combined blur from scope aperture and mount tracking errors will be greater than 1.6" FWHM stars.

If you are so set at 533 sensor - by all means go for it. And if you don't want to get reducer flattener as well (533 won't need it as it is 1" sensor), that is fine as well. In that case, do seriously consider how you are going to process your data. You'll need to bin x2 at least (and binning of OSC data is not straight forward as with mono - if you debayer with interpolation and then bin - you don't get same SNR improvement as most of your pixels have correlation).

An excellent example Vlaiv, and an explanation to go with it, thank you. I have learn't in a very short space of time, to respect your knowledge. Astro imaging is now just something I want to 'dabble' with, while obtaining some reasonable results, hence opting for a Tripet, as CA would be very much a no-go for me. I have experienced it with Doublets.  Back in my days of astro imaging, I had many scopes, 8, 10, and 12" Meade SCT,  and various refractors, even a Coronado Solar Scope. My last set-up was an 8" Meade SCT mounted alongside a WO FLT 110, and SX CCD Cameras. All permanently mounted on an NEQ6 in an observatory. As you read in my into, this all came to end in 2014, following a serious medical issue. Now I am just wanting to "scratch an itch" than has never gone away. My prime interest now is photography, and for that I use all 'Pro Spec' gear.

Link to comment
Share on other sites

A question then Vlaiv, if I may. Given that the starburst and any amp glow can potentially be removed with Dark Frames, should  I be considering pixel size (3.7um with the 533, and 4.63 with 294), as opposed to Glow/Starburst. That is 0.61" secs per pix for the 533, and 1.19 arc secs per pix for the 294.  The 294 has 11.7MP as opposed to the 9Mp of the 533, so effectively more pixels from which to bin. Also, a little more forgiving on the guiding/tracking.  There is only 100GBP difference in the price between the two cameras, the 294 being the more expensive.

Link to comment
Share on other sites

3 hours ago, centroid said:

A question then Vlaiv, if I may. Given that the starburst and any amp glow can potentially be removed with Dark Frames, should  I be considering pixel size (3.7um with the 533, and 4.63 with 294), as opposed to Glow/Starburst. That is 0.61" secs per pix for the 533, and 1.19 arc secs per pix for the 294.  The 294 has 11.7MP as opposed to the 9Mp of the 533, so effectively more pixels from which to bin. Also, a little more forgiving on the guiding/tracking.  There is only 100GBP difference in the price between the two cameras, the 294 being the more expensive.

I've found that amp glow / starburst thing is properly calibrated out on all my CMOS sensors - as long as they are thermally stable, and I don't see that being an issue at all.

In fact - here is stretched master dark for above image of M51:

image.png.b78c58f94c5cde58dea7396b114b831d.png

While not quite starburst kind of glow - it is still present in stretched dark (this one being binned x8 for size/display)

Here is master dark from my other camera:

image.png.fc40603164795b0236a26bc8d01ab486.png

This one is real starburst (ASI178mcc - cooled color model, now discontinued)  - yet it also calibrates out.

Important thing to note with these CMOS sensors - is to skip bias - it is not needed, to calibrate with darks of exact exposure length (no dark scaling unfortunately - if that is important to you).

As far as pixel size / sensor size thing, here is my advice and reasoning behind it. Get largest affordable sensor that has set point cooling :D. Sensor size is speed. I see a lot of people looking for "fast" scopes - in terms of F/ratio, but F/ratio is not equal to speed of telescope (here speed refers to time needed to get to target SNR or alternatively - SNR achieved in set amount of time).

F/8 telescope can be faster than F/5 telescope - if paired with proper camera / pixel size.

Take for example this:

image.png.f1b2fa76e2c02c4f2ac04adf32969975.png

FOV is comparable in terms of size, one is F/9 scope and other is F/7 scope. If one bins pixels to get to same sampling rate (arc seconds per pixel) - it is the aperture that rules. 6" of aperture will gather more light than 4.5" - simple as that (and of course - made possible by larger sensor).

For that reason, I advocate larger sensor - 294 even if there is amp glow / starburst - it can simply image more of the sky in single go (want to take larger target with smaller sensor? - need to do mosaics - with larger sensor it might fit in FOV at once. Only restriction is fully corrected field diameter of telescope).

In fact - between 533 and 183 - I'd choose 183. They are of approximately the same size, and while 533 is newer sensor and free from amp glow - I advocate full / proper calibration anyway and amp glow is there fore non issue really - ASI183 has smaller pixel - which means more flexibility in the way you bin / debayer your data.

With 2.4µm pixel size - 805mm of focal length gives 0.61"/px - that is good "baseline" to be binned as bin x2 will give you 1.22"/px, Bin x3 will be 1.83"/px and Bin x4 will give you 2.44"/px. I believe that last two would be most used resolutions with 115mm scope - 1.83"/px and 2.44"/px

 

  • Thanks 1
Link to comment
Share on other sites

Posted (edited)
2 hours ago, vlaiv said:

I've found that amp glow / starburst thing is properly calibrated out on all my CMOS sensors - as long as they are thermally stable, and I don't see that being an issue at all.

In fact - here is stretched master dark for above image of M51:

image.png.b78c58f94c5cde58dea7396b114b831d.png

While not quite starburst kind of glow - it is still present in stretched dark (this one being binned x8 for size/display)

Here is master dark from my other camera:

image.png.fc40603164795b0236a26bc8d01ab486.png

This one is real starburst (ASI178mcc - cooled color model, now discontinued)  - yet it also calibrates out.

Important thing to note with these CMOS sensors - is to skip bias - it is not needed, to calibrate with darks of exact exposure length (no dark scaling unfortunately - if that is important to you).

As far as pixel size / sensor size thing, here is my advice and reasoning behind it. Get largest affordable sensor that has set point cooling :D. Sensor size is speed. I see a lot of people looking for "fast" scopes - in terms of F/ratio, but F/ratio is not equal to speed of telescope (here speed refers to time needed to get to target SNR or alternatively - SNR achieved in set amount of time).

F/8 telescope can be faster than F/5 telescope - if paired with proper camera / pixel size.

Take for example this:

image.png.f1b2fa76e2c02c4f2ac04adf32969975.png

FOV is comparable in terms of size, one is F/9 scope and other is F/7 scope. If one bins pixels to get to same sampling rate (arc seconds per pixel) - it is the aperture that rules. 6" of aperture will gather more light than 4.5" - simple as that (and of course - made possible by larger sensor).

For that reason, I advocate larger sensor - 294 even if there is amp glow / starburst - it can simply image more of the sky in single go (want to take larger target with smaller sensor? - need to do mosaics - with larger sensor it might fit in FOV at once. Only restriction is fully corrected field diameter of telescope).

In fact - between 533 and 183 - I'd choose 183. They are of approximately the same size, and while 533 is newer sensor and free from amp glow - I advocate full / proper calibration anyway and amp glow is there fore non issue really - ASI183 has smaller pixel - which means more flexibility in the way you bin / debayer your data.

With 2.4µm pixel size - 805mm of focal length gives 0.61"/px - that is good "baseline" to be binned as bin x2 will give you 1.22"/px, Bin x3 will be 1.83"/px and Bin x4 will give you 2.44"/px. I believe that last two would be most used resolutions with 115mm scope - 1.83"/px and 2.44"/px

 

Again, thanks Vlaiv, comprehensive, and informative👍. The time and effort taken to compile your responses, is much appreciated. I think I will go for the larger 'chip' and pixel IMX 294, and deal with any glow with dark frame calibration. I will be  going for the 'cooled' version camera, which should allow for identical darks. The HEQ5 Pro has been shipped, and should arrive tomorrow.

Edited by centroid
Link to comment
Share on other sites

  • 3 weeks later...
On 16/08/2021 at 17:11, centroid said:

115mm f/7 APO refractor, on an HEQ5-PRO mount.

Sony IMX533 or Sony IMX183

I think that the 294 would be a better choice with that scope. 

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.